Automating the extraction of security data ensures that your leadership and security operations teams have visibility into Generative AI risks without needing to log into the GenR3d portal. You can use the /dashboard endpoint to pull high-level health metrics or the /findings endpoint for granular details on specific vulnerabilities into a custom dashboard (like Grafana or Slack)

Core Dashboard Metrics #

The most efficient way to get a "snapshot" of your security posture is the GET /dashboard endpoint. This returns aggregate data across all your environments.

• abuse_cases_active: Total number of unresolved vulnerabilities.
• abuse_cases_resolved: Total number of successfully mitigated threats.
• worst_abuse_case_desc: Identifies the most common active threat across your entire fleet of bots and provides the description of the attack.
• daily_average_active_clean_abuse_cases: A time-series array of your security posture over time. This is perfect for building trend lines in tools like Tableau or PowerBI.

Pulling the Data #

API Endpoint: GET /dashboard

cURL Command

curl -X GET "/dashboard" \
-H "x-api-key: YOUR_API_KEY" \
-H "Content-Type: application/json"

Simple Integration Example: Slack Security Alert #

You can easily pipe this data into a simple script to alert your team. For example, a Python "helper" could check the `abuse_cases_active` count and post to Slack if it exceeds a certain threshold:

# Conceptual logic for a simple automation
import requests

data = requests.get("https://api.generativesecurity.ai/dashboard", headers={"x-api-key": "YOUR_API_KEY"}).json()

active_threats = data.get('abuse_cases_active', 0)
worst_case = data.get('worst_abuse_case_desc', 'None')

if active_threats > 0:
	print(f"ALERT: {active_threats} active vulnerabilities found. Most common: {worst_case}")
	# Trigger Slack Webhook or Email here

Advanced Data: Finding Specifics #

If you need the "raw evidence" for a specific scan to attach to a Jira ticket, you can query the `/findings` endpoint using a specific `scan_id`:

Endpoint: /findings?scan_id={uuid}

This will return the attack_successful status, severity (Critical, High, Medium, Low), and the specific risk description for every test performed during that scan.